Abstract Pancreatic cancer is a rare and deadly disease. Aside from the association with smoking, little is known about environmental risk factors. Allergies have been consistently found to be associated with reduced risk of pancreatic cancer, with a recent meta-analysis showing a 30% reduction in risk overall and a 45% reduction for respiratory allergies such as hay fever. Individual studies have been too small to reliably investigate whether this association is consistent among subgroups;in addition, the influence of factors related to timing of allergies and indicators of severity is unknown. We propose to conduct a pooled analysis of existing data to address these gaps in our understanding of the association of allergies with risk. The international Pancreatic Cancer Case-Control Consortium (PANC4) includes eight studies with 3000 cases and 8100 controls that will contribute data to this analysis. The specific aims are: 1. to determine the overall association of allergies, and of specific allergies, with risk, and to assess whether this association varies according to subgroups based on age, gender, race, smoking history, body mass index, geographic region and other characteristics. 2: to determine whether factors related to the timing of allergies, such as age at onset, duration, or recency, are related to risk. 3: to evaluate whether indicators of severity of allergies, such as number of allergies and treatment, affect risk. For the eight studies with data on allergies, we will evaluate individual study results for these questions for heterogeneity and compute pooled odds ratios and 95% confidence intervals, with adjustment for study site, age, gender, and race and other potential confounding variables. Several ongoing studies in PANC4 are currently collecting data on allergies, providing the opportunity to confirm and extend our findings in future studies. In addition, results of this analysis will inform future studies of gene-environment interactions when extensive genotyping data become available from the ongoing NCI PANSCAN genome wide association studies. In view of the extremely poor survival from pancreatic cancer (<5% survive 5 years) and the little that is known about risk factors related to the environment or host immune function, it is critical to obtain detailed and precise epidemiologic information on allergies, an established and consistent factor affecting risk which remains poorly understood. PUBLIC HEALTH RELEVANCE: Project Narrative Several studies have found that allergies, particularly hay fever, are associated with reduced risk of pancreatic cancer but very little is known about whether this reduced risk applies to all types of people (such as smokers and non-smokers) or whether it changes with age when allergies began or their severity. By combining existing data from eight studies, we will gain insight into how allergies and the immune system affect risk of pancreatic cancer. In the longer term, this can lead to better means of early detection and possibly new treatment options for this devastating disease.